active
Quantitative and machine-learning fractography
Using surface morphology, computer vision, and statistical descriptors to turn fracture surfaces into mechanical evidence.
Fractography is treated here as a quantitative data source rather than only a post-failure image record. Surface morphology can preserve information about failure processes, loading conditions, and material resistance.
This direction includes roughness/toughness correlations, correlation length scales, computer vision, machine learning, and interpretable descriptors for fracture surfaces.